At the Edge: Exploratory Spatial Data Analysis and Terrorism Research

• In Event: Poster Session: State Politics & Conflict Processes: Violence and Contention by States and Non-State Actors
  In Poster Session: Poster Session: Conflict Processes: Violence and Contention by States and Non-State Actors

Abstract

Current empirical models for understanding the causes of terrorism are limited in their capabilities for explaining causes at the local level. Researchers often make broad assertions based on large datasets without assessing geographic variation in terrorism, or researchers focus on groups as a level of analysis. However, the timing and targets of terrorism cannot be understood without assessing how terrorism varies across time and space. Exploratory spatial data analysis (ESDA) allows researchers to explore the spatial (and, in some cases, temporal or spatiotemporal) patterns of terrorist attacks in different environments. ESDA allows the researcher to describe spatial distributions, identify spatial outliers, and look for regionalization (clustering and clusters) of terrorist attacks. An analysis of the spatial distribution of terrorism allows researchers to understand better where terrorism occurs and the geographic features (population, landmarks, etc.) associated with terrorism. While the analysis of spatial outliers and clustering shifts the unit of analysis for research to clusters. This shift aligns methodologically with the extant literature’s hypotheses that imply that attacks cluster (e.g., outbidding, electoral violence, response to changes in the local context) and that groups participate in “campaigns” of terrorism.

Using a newly compiled and georeferenced dataset on terrorism in Northern Ireland, I explore the utility of ESDA as a tool for understanding the targets and timing of terrorism. Using spatial point pattern analysis, parallel coordinate plots, spatial autocorrelation, and other exploratory spatial methods, I look at the timing and targets of terrorism in Northern Ireland between 1996 and 2001. Exploratory analysis allows me to identify interesting spatial, temporal, and spatiotemporal patterns in terrorism data. As with other exploratory analyses, ESDA identifies patterns that could result from multiple processes. As such, this type of analysis allows for the identification of patterns and locations of interest that require analysis with other methodologies. Exploratory spatial data analyses are largely absent from the existing research, and with this paper, I offer both examples and suggestions for its integration into the study of terrorism.

Author

• Shannon Colleen Hartmann, University of Nevada, Reno